P‐3.1: Navigating Privacy Concerns: A Comprehensive Review of Challenges in Artificial Intelligence Image‐to‐Video Generation Applications

Artificial Intelligence Generated Content (AIGC) has gained significant popularity in recent years, particularly in the domain of video generation. Among its key branches, Image‐to‐Video AI focuses on transforming static images into dynamic video content through innovative techniques. Leveraging advanced artificial intelligence methods, especially deep learning and generative models, this technology enables the creation of coherent video sequences from one or more input images. Image‐to‐Video AI has rapidly captured a growing share of the video creation industry, with frequent updates and iterations in technical architectures. However, this rapid advancement has introduced substantial privacy risks and challenges, including but not limited to issues of data source legality, risks of privacy infringement in generated content, susceptibility to adversarial attacks and privacy inference, and risks of unauthorized dissemination. This paper explores the technological advancements in Image‐to‐Video AI, providing insights into the privacy challenges associated with its applications and implementation. Furthermore, it advocates for future research efforts to address these challenges, ensuring the secure and effective deployment of Image‐to‐Video AI technologies. Emphasis is placed on fostering industry standardization, enabling privacy‐preserving data collection and content creation, and developing robust post‐deployment safeguards.